healing the wounds: learning from sierra leone's post-war...
TRANSCRIPT
NBER WORKING PAPER SERIES
HEALING THE WOUNDS:LEARNING FROM SIERRA LEONE'S POST-WAR INSTITUTIONAL REFORMS
Katherine CaseyRachel Glennerster
Edward Miguel
Working Paper 18368http://www.nber.org/papers/w18368
NATIONAL BUREAU OF ECONOMIC RESEARCH1050 Massachusetts Avenue
Cambridge, MA 02138September 2012
This paper draws together findings from a number of different research projects. The authors gratefullyacknowledge generous financial support from the National Bureau of Economic Research AfricanSuccesses Project, the GoBifo Project, the Harry F. Guggenheim Foundation, the Institutional Reformand Capacity Building Project (IRCBP), the International Growth Centre (IGC), the International Initiativefor Impact Evaluation (3ie), the MIT Horace W. Goldsmith Foundation and the World Bank DevelopmentImpact Evaluation (DIME) Initiative. The views expressed herein are those of the authors and do notnecessarily reflect the views of the National Bureau of Economic Research.
NBER working papers are circulated for discussion and comment purposes. They have not been peer-reviewed or been subject to the review by the NBER Board of Directors that accompanies officialNBER publications.
© 2012 by Katherine Casey, Rachel Glennerster, and Edward Miguel. All rights reserved. Short sectionsof text, not to exceed two paragraphs, may be quoted without explicit permission provided that fullcredit, including © notice, is given to the source.
Healing the Wounds: Learning from Sierra Leone's Post-war Institutional ReformsKatherine Casey, Rachel Glennerster, and Edward MiguelNBER Working Paper No. 18368September 2012JEL No. F35,H41,O40
ABSTRACT
While its recent history of civil war, chronic poverty and corrupt governance would cause many todismiss Sierra Leone as a hopeless case, the country's economic and political performance over thelast decade has defied expectations. We examine how several factors—including the legacy of war,ethnic diversity, decentralization and community-driven development (CDD)—have shaped localinstitutions and national political dynamics. The story that emerges is a nuanced one: war does notnecessarily destroy the capacity for local collective action; ethnicity affects residential choice, butdoes not impede local public goods provision; while politics remain heavily ethnic, voters are willingto cross ethnic boundaries when they have better information about candidates; decentralization canwork even where capacity is limited, although the results are mixed; and for all of its promise, CDDdoes not appear to transform local institutions nor social norms. All of these findings are somewhat“unexpected,” but they are quite positive in signaling that even one of the world’s poorest, most violentand ethnically diverse societies can overcome major challenges and progress towards meaningful economicand political development.
Katherine CaseyStanford Graduate School of Business 655 Knight WayStanford, CA [email protected]
Rachel GlennersterAbdul Latif Jameel Poverty Action LabMIT Department of EconomicsE60-275Cambridge MA [email protected]
Edward MiguelDepartment of EconomicsUniversity of California, Berkeley530 Evans Hall #3880Berkeley, CA 94720and [email protected]
1
1. Introduction
Academics and development practitioners agree that strengthening the transparency,
accountability and inclusiveness of institutions could be important determinants of economic
performance (Engerman and Sokoloff 1997, Acemoglu, Johnson and Robinson 2001, Banerjee
and Iyer 2004). Yet they also acknowledge that it remains unclear what types of interventions
could successfully make progress towards these objectives. Such concerns take on heightened
urgency in a country like Sierra Leone, which has suffered decades of extreme poverty and
recently emerged from a devastating civil war. In this context, the objectives of inclusive and
participatory governance are twofold. First, enhancing accountability aims to provide a more
effective vehicle to channel government and donor resources towards the reconstruction of
public infrastructure and restoration of basic services. Second, creating avenues for public
participation allows citizens to voice and seek redress for grievances regarding government
incompetence and corruption, as well as confront and amend long standing social tensions and
inequities that many believed helped fuel the recent violence.
This paper examines how different factors—including the legacy of war, ethnic diversity,
decentralization and community-driven development (CDD)—affect local institutions and
collective action, as well as national political culture and outcomes in Sierra Leone. The story
that emerges is nuanced and does not confirm reflexive biases: war does not necessarily destroy
the capacity for local collective action; ethnicity affects residential choice, but does not impede
local public goods provision; while politics remain heavily ethnic, voters are more willing to
cross ethnic boundaries in local elections where they have better information about candidates;
decentralization can work even where local capacity is highly constrained, although the results
are mixed; and for all of its promise and some positive impacts on local public goods provision,
CDD does not transform local institutions nor social norms of behavior. All of these results are
somewhat “unexpected,” but they are quite positive in signaling that even one of the world’s
poorest, most violent and ethnically diverse societies can overcome major challenges and
progress towards meaningful economic and political development.1
1 This paper summarizes and brings together findings from four others: Bellows and Miguel (2009); Glennerster, Miguel and Rothenberg (2010); Casey, Glennerster and Miguel (forthcoming); and Casey (2012). For complete details on theoretical models, identification strategies, empirical specifications and datasets, please see these original papers.
2
The rest of the paper is structured as follows. Section 2 provides background on Sierra Leone’s
protracted decline into poverty and unrest, exploring prominent social divisions that may have
encouraged young men to take up arms. Section 3 discusses the historical evolution of ethnic
diversity and the complex role it plays in contemporary public life. Section 4 details key post-
war institutional reforms including the restoration of multiparty democracy, decentralization and
CDD; while Section 5 assesses the progress these reforms have mode toward encouraging
economic development, democratizing institutions and changing social norms. Section 6
concludes.
2. Legacies of Poverty, Corruption and Conflict
After achieving independence from Britain in 1961, Sierra Leone enjoyed only a brief period of
free and competitive democracy. Increasing political instability, worsening governance and
deepening poverty marred the subsequent few decades, which terminated in institutional collapse
and civil war. In the 1970’s and 1980’s the country was ruled by authoritarian leaders who
enriched themselves through illicit deals involving diamonds, while doing little to provide
needed services such as health care and education (Reno 1995). Eliminating threats to its
absolute control, the government of President Siaka Stevens dismantled competitive democracy
by abolishing district-level local government in 1972 and declaring the country a one party state
in 1978. By the early 1990’s, Sierra Leone had the second lowest living standards of any
country in the world (United Nations 1993).
Figure 1 situates the economic experience of Sierra Leone in relation to the rest of sub-Saharan
Africa. The comparative poverty of Sierra Leone is evidenced by real GDP per capita that falls
significantly below the regional average for the entire period (1960-2010). Note further the
economy’s stagnant performance from 1970-90, followed by precipitous decline during the war.
This paper focuses on the institutional factors surrounding the positive post-war recovery
apparent in the upward trend of the last decade. While the particular confluence of events and
reforms is unique to Sierra Leone, note that its strong performance in the last several years
mimics broader trends of growth for the region as a whole (Miguel 2009).
3
The weak economic performance and poor governance of the 1970’s and 80’s steered the
country toward civil unrest. Partially as a result of the widespread discontent towards the
corruption and ineffectiveness of the government, a small group of rebels, who had entered the
country from Liberia in 1991, were successful in recruiting disenfranchised youth to rise up
violently against the status quo. As their numbers swelled by early 1992, these rebels, known as
the Revolutionary United Front (RUF), spread the armed conflict to all parts of the country. The
brutal civil war that ensued saw an estimated 50,000 Sierra Leoneans killed, over half of the
population displaced from their homes, and thousands of civilians victimized by amputation,
rape, and assault (Human Rights Watch 1999). A small cadre of British troops, along with a
large international peacekeeping mission, brought the war to a decisive end with peace officially
declared in January 2002.
Scholars point to a number of long standing social divisions that created frustration and may
have helped incite violence. First, some have claimed that the initial motivations of the RUF
were idealistic and that the early rebels were guided by a strong sense of political grievances
related to the failings of the corrupt regime (Richards 1996). Such frustrations were particularly
acute for young men, who were largely excluded from decision making and at times subject to
coerced labor and capricious fines by traditional authorities. Second, colonial rule enhanced the
historical legacy of inequality between local chiefs and their subjects. During the colonial
period, the British implemented direct rule over residents of the Western peninsula (the Colony),
yet exerted indirect rule over residents of the interior (the Protectorate). This latter system
promoted chiefs loyal to the British, and institutionalized – and in many cases augmented – their
autocratic power over their subjects, thereby exacerbating inequality and reinforcing social
divisions. Third, although not likely a direct cause of violence, women have historically held
less power in local governance and possessed weaker socioeconomic status as compared to men.
After the war ended, major institutional reforms aimed to address these root causes of dissension,
to both promote greater equity and preclude a return to violence.
Although devastating, the war did not leave the country so weakened as to be incapable of
recovery. While violence inflicted during the war created incalculable human suffering and
4
destroyed much of the physical infrastructure of the country, the impact on institutions is more
nuanced. Bellows and Miguel (2009) unexpectedly uncover a positive association between
exposure to violence and subsequent increases in political and social activism. Their research
suggests that individuals whose households directly experienced war violence are more active
political and civic participants than non-victims: they are more likely to vote (by 2.6 percentage
points), attend community meetings (by 6.5), belong to a social (by 6.6) or political (by 5.7)
group, and serve on school management committees (by 3.8). In addition, these victims were no
worse off in terms of standard consumption measures a few years after the war ended. While
these findings underscore the extraordinary resilience of Sierra Leoneans, it is important to note
that they are based on variation across individuals within the same village, and thereby do not
estimate the net effect of civil war on the country as a whole. The authors conclude that while
the “humanitarian costs of civil wars are horrific… it appears their legacies need not be
catastrophic.”
3. Ethnic Diversity
Ethnicity plays a nuanced role in the social and political life of Sierra Leone, defying commonly
held conceptions about the adverse effects of diversity. Many scholars have argued that ethnic
diversity is an important impediment to economic and political development. Economic growth
rates are slower in ethnically diverse societies, and local public goods provision often suffers
(Easterly and Levine 1997, Alesina et al. 1999, Alesina et al. 2003, Fearon 2003). The inability
to overcome the public good free-rider problem in diverse communities, due to monitoring and
enforcement limitations, is the leading explanation proposed for less developed countries
(Miguel and Gugerty 2004, Habyarimana et al 2007, 2009). These issues are particularly salient
in sub-Saharan Africa, the world’s most ethno-linguistically diverse region. Yet our research
shows that while ethnicity is important for residential choice and political allegiances in Sierra
Leone, it does not appear to hamper local collective action and was not an organizing factor in
the civil war.
5
3.1 Historical View of Ethnicity
As background, Sierra Leone is very diverse, ranking fifteenth on the Taylor and Hudson (1972)
list of countries with the highest levels of ethno-linguistic fractionalization.2 Specifically, of
eighteen major ethnic groups, the Mende and Temne are numerically dominant, occupying
shares of 32.2% and 31.8%, respectively; while the Limba, Kono, and Kuranko are the next
largest groups, at 8.3%, 4.4%, and 4.1%, respectively (National Population and Housing Census
2004). Other groups occupy a substantially smaller share, including the Krio (Creoles)—former
slaves who returned to Africa to settle Freetown—whose population share fell to only 1.4% by
2004. These groups are characterized by distinct customs, rituals, and history, and, most
importantly, language. With the exception of Krio, an English dialect, the other languages are
members of the Niger-Congo language family. Within this family, the most salient distinction is
between the Mande languages—including Mende, Kono, Kuranko, Susu, Loko, Madingo,
Yalunka, and Vai—and the Atlantic-Congo languages, including Temne, Limba, Sherbro,
Fullah, Kissi, and Krim. These groups are mutually unintelligible to each other, and much further
apart linguistically, for example, than English and German.
Over the past two centuries national politics have been heavily influenced by two distinct
divisions along ethnic lines, where the early Colony versus Protectorate tension under the British
was later surpassed by regional allegiances after independence. At the time of the founding of
the Sierra Leone colony in the late 18th century and through much of the 19th century, the Krio
enjoyed a relatively privileged political and economic position due to their facility with English
and special links with the British even though they were numerically small. Before
independence, the key political division in Sierra Leone was Krio vs. non-Krio, but because of
growing tensions between the Krio and “up country” ethnic groups, the British progressively
limited their political power. Thus as Sierra Leone made its transition to independence, the
primary source of political conflict shifted. As stated by Kandeh (1992), “the salience of the
Creole [Krio]-protectorate cleavage was eclipsed after independence by the rivalry between the
Mendes of the south and Temnes of the north.” It is this largely regional divide that continues to
galvanize national politics today.
2 As cited in Easterly and Levine (1997).
6
Two facts about the Krio may have prevented ethnic political divisions from escalating into
violent conflict. One key difference between Sierra Leone and many other African countries is
that the “favored” ethnic group during early colonialism was not truly indigenous and no longer
holds a position of power. They historically served as a common antagonist for the Mendes and
Temnes together, and have since lost their political influence. Second, the Krio people gave
Sierra Leone their language, also called Krio, which is a dialect of English that has been
influenced by Portuguese, Arabic, Yoruba and many African languages as a legacy of the slave
trade. Serving as a national lingua franca for decades, Krio is currently spoken (usually as a
second language) by most Sierra Leoneans, and is increasingly taught in schools. In many other
African countries the lingua franca is the former colonial language, usually English or French.
While Krio has a base in English, it is unique to Sierra Leone and widely spoken even by those
with no schooling. While the existence of a common national language is clearly insufficient to
guarantee social stability—as the African cases of Rwanda and Somalia poignantly illustrate—
Krio’s ubiquity in Sierra Leone may (through historical accident) help promote the consolidation
of a common national identity that transcends tribe (Ngugi 2009), as with Swahili in post-
independence Tanzania (Miguel 2004).
Perhaps due to these unifying forces and in contrast to most popular media coverage on African
civil wars, neither ethnic nor religious divisions played a central role in the Sierra Leone conflict.
The Revolutionary United Front (RUF) rebels targeted people from all ethnic groups, and
statistical analysis of documented human rights violations shows that no ethnic group was
disproportionately victimized. There is also no evidence that civilian abuse was worse when
armed factions and communities belonged to different ethnic groups (Humphries and Weinstein
2006). Ethnic grievances were not rallying cries during the war and all major fighting sides were
explicitly multi-ethnic (Keen 2005). The fact that the war was not fought along ethnic lines, and
the central role that external actors played in bringing it to a conclusive end, may partially
explain why there has not been a resurgence of violence.
3.2 How Ethnicity Matters Today
The fact that ethnic identity was not an organizing factor in the conflict does not, however, mean
that it is unimportant. Glennerster, Miguel and Rothenberg (2010) show that a preference for
7
one’s own ethnic group is a key determinant of residential choice for rural Sierra Leoneans.
Using nationally representative household data that collected respondents’ chiefdom of residence
in 1990 and 2007, they estimate a discrete choice model to understand why different individuals
moved across chiefdoms after the war. Since contemporary ethnic composition is in part
endogenously determined by post-war migration choices, their empirical specifications use
historical 1963 ethnic shares. They find that individuals are on average willing to travel an
additional 10.1 kilometers to live in a chiefdom with a 10 percentage point greater share of
her/his own ethnic group. While still strong, the co-ethnic preference is attenuated for people
with some education (note that adult literacy is just 34.8%). In particular, educated individuals
are only willing to travel an additional 8.6 kilometers to live in a chiefdom with 10 percentage
point greater share of her/his own ethnic group, which suggests that education dampens co-ethnic
residential preferences.
Yet the preference for living with one’s own ethnic group does not translate into a weaker ability
to work together with those from other groups. In fact, conditional on other factors (including
remoteness from cities as well as population size and density), individuals exhibit a positive
preference for diversity, though this is smaller than the preference for a higher co-ethnic share.
Standing in sharp contrast to the bulk of the ethnic diversity literature, Glennerster et al. extend
their analysis and find no negatives effects of ethnic diversity on the provision of local public
goods. The authors use a mean effects approach to summarize the average impacts of the project
across a family of related indicators (following Kling et al. 2007). Specifically, they find no
effect of ethnic—or religious—diversity on local collective action (as measured by road
maintenance, group membership, self-expressed trust or disputes); no effect of ethnic diversity
on the quality of primary schools (as measured by instructional supplies, facilities or teaching);
and if anything, positive impacts of ethnic diversity on health clinic quality, supplies, and staff
presence and quality.
Arguably the most salient arena of ethnic loyalties today lies in politics. Casey (2012) notes the
tight correlation between voting choices and ethnic identity. The two major political parties—
the Sierra Leone People’s Party (SLPP) and the All People’s Congress (APC)—have strong,
long-standing ties to the Mende and other ethnic groups in the South and the Temne and other
8
groups in the North, respectively. As an example of the strength of these loyalties, in the 2007
Parliamentary elections the APC won 36 of 39 seats in the Northern Province, while the SLPP
and its splinter party, the People’s Movement for Democratic Change (PMDC), swept 24 of 25
seats in the South. At the individual level, voters are readily forthcoming about ethnic-party
allegiances: laughter is not an uncommon response to questions about why someone voted for a
particular political party, often followed by an explanation of how their father and grandfather
voted for the same one. When asked which party they supported in the first round of the 2007
Presidential elections, 88.4% of Mende voters reported choosing the SLPP or PMDC, while
94.2% of Temne voters reported backing the APC.3 Yet Casey argues that these traditional
ethnic alliances are not immutable, particularly in local elections where voters have access to
better information about candidates. Detailed exit poll survey data collected in 2008 shows that
while party is equally as important as a bundle of individual candidate characteristics (like
reputation in his/her previous job, education and kinship ties) in national elections, voters say
that party is only half as important as these traits in local races.
4. Institutional Reforms
Sierra Leone emerged from the war in a very weak position economically, socially and
politically. The country remained at the bottom of the UN Human Development Index ranking;
families were mourning the loss of loved ones and grappling with the emotional and physical
traumas of war; the network of infrastructure and public services was largely destroyed; and the
government faced an institutional vacuum. At the same time, the people were ready for change
and committed to preventing a return to violence. The Government of Sierra Leone and its
donor partners responded with an ambitious program of institutional reform and economic
development. The reforms aimed to foster economic growth while strengthening the institutional
environment to promote better governance and lasting peace.
3 Source: IRCBP, National Public Services Survey 2008. This is a nationally representative household survey of over 6,000 households. Survey responses are limited to those respondents who stated their voting choice and could verify the fact that they had voted by producing their voter identification card with a hole punch made by polling center staff indicating participation in the 2007 first round voting.
9
The most high profile political reforms were the restoration of multiparty democracy and the
reconstitution of local government after over 30 years of dormancy. The National Electoral
Commission (NEC) oversaw competitive elections for Parliament and the Presidency in 2007
and for the 19 Local Councils in 2004 and 2008. These reforms brought democratic competition
back to national, district and ward-level institutions. The peaceful transfer of power from the
ruling SLPP regime to the opposition APC challengers in 2007 stands out as a particularly
impressive achievement. When compared to its counterpart for the Kenyan elections of the same
year, the NEC distinguished itself by taking decisive action to address charges of electoral
irregularities from both sides—the challengers and the ruling party. A report by Freedom House,
the NGO watchdog, notes how the NEC “functioned with remarkable independence and helped
to ensure the success of the balloting, despite postponements and other difficulties.”
Understanding the factors that enable electoral commissions to succeed—and what role external
donors might play in helping to protect their independence (the UN among others provided
substantial support in Sierra Leone)—is important.
In addition to historical antecedent, the main theoretical motivations for decentralization seemed
a good match for the post-war context, both in terms of providing local public goods and
empowering the citizenry. On the public goods front, theorists point to economic welfare gains
from decentralizing service provision to local governments that have better information about,
and greater ability to tailor outputs to, differences in local preferences and costs (Oates 1999).
By reducing the (geographic and bureaucratic) distance between frontline service providers and
managers, decentralization can further reduce the cost of supervision and increase the speed and
efficiency with which managers respond to needs on the ground. The informational and
supervisory advantages of local government are particularly important in light of weak
transportation and communication networks that exacerbate central government oversight
challenges. Yet it is not clear that decentralized provision necessarily dominates centralized, as
there are risks that the accountability gains from better information may be compromised by a
greater risk of elite capture (particularly when only a small minority is literate and politically
aware) or that services management will suffer in the hands of less competent local politicians
(Bardhan and Mookerjee 2006).
10
Aware of the potential benefits and risks, the Government of Sierra Leone began to gradually
transfer responsibilities and tied grants for public services in health, agriculture, education and
other sectors over to the Local Councils in accordance with the Local Government Act of 2004.
The extent of decentralization to date varies widely across sectors, where health takes the lead
and education lags conspicuously behind. Yet even in health, decentralization remains at best
partial since the power to hire, fire and remunerate staff is retained by the central government.
To complement the conditional grants by sector, the Councils were given several million dollars
worth of discretionary funds to use toward development projects in their districts under the Local
Government Development Grants (LGDG) program.
Turning to civic engagement, scholars of decentralization emphasize the political value of
creating greater opportunities for citizens to participate in government, which can be particularly
important in developing countries that inherited highly centralized regimes from colonial powers
(Oates 1999). They suggest that inclusion and participation carry both intrinsic value in
empowering citizens as well as accountability gains in enabling the public to better monitor and
constrain the behavior of elected officials. In Sierra Leone encouraging the participation of
women and youths takes on added importance in light of their long standing exclusion and the
role that resulting frustrations may have played in the war.
The government and donors sought to complement these national reforms with “bottom-up”
initiatives aimed at strengthening community level institutions. Community driven development
(CDD) is one such approach that has become very popular throughout the developing world: the
World Bank alone has spent US$50 billion on CDD initiatives over the past decade (Mansuri and
Rao 2012). By emphasizing local participation in and control over project implementation, CDD
aims to provide public goods through a process that empowers the poor. Advocates of
participatory local governance promise a long and varied list of benefits ranging from more cost-
effective construction of infrastructure, to a closer match between project choice and village
needs, to the weakening of authoritarian village institutions.4 Critics hold concomitant concerns
4 For instance, Dongier et al. (2003) write that: “Experience demonstrates that by directly relying on poor people to drive development activities, CDD has the potential to make poverty reduction efforts more responsive to demands, more inclusive, more sustainable, and more cost-effective than traditional centrally led programs…achieving immediate and lasting results at the grassroots level.”
11
that participation requirements serve as a regressive tax, widening political participation clogs up
rather than expedites decision-making (Olson 1982), and external resources attract new leaders,
crowd out the most disadvantaged (Gugerty and Kremer 2008) or are captured by elites if the
program is unable to change the nature of de facto political power (Bardhan 2002). Any real
world program risks manipulation during implementation, and skeptical observers fear that
donors simply use the jargon of participatory development for political or public relations
purposes while continuing to operate in a “top-down” manner. Few studies provide rigorous
empirical evidence regarding these claims (Mansuri and Rao 2004).
Using a randomized experiment, Casey, Glennerster and Miguel (forthcoming) evaluate the
success of one particular CDD project in Sierra Leone, the “GoBifo” project (which means
“Move Forward” in Krio). As is typical of CDD projects, GoBifo established village-level
structures and provided tools to plan and manage development resources; provided communities
with financing and technical assistance in implementing small scale projects (totaling $4,667 per
community or roughly $100 per household); and created links between these processes and local
government institutions. This “hardware” support was coupled with intensive community
facilitation, or “software,” that promotes democratic decision-making, the participation of
socially marginalized groups, and transparent budgeting practices. The CDD approach attempts
to bolster local coordination—for example, by setting up village development committees—and
to enhance participation, by requiring women and “youths” (adults under age 35) to hold
leadership positions, sign off on project finances, and attend meetings. Project emphasis on
inclusive and democratic decision-making aims to empower women and youth in other realms of
local governance; and the learning-by-doing experience with successfully implementing
communal projects aims to catalyze collective action beyond the immediate project sphere. The
idea is that once communities have the institutions in place—a village development committee,
plan, bank account and experience in budgeting and management—they should be better able to
take advantage of new opportunities that arise after the program itself has ended. This latter
emphasis on “help yourself” activism echoes current President Ernest Bai Koroma’s “Attitudinal
Change” public messaging campaign that urges citizens to take responsibility for their own
development.
12
5. Effectiveness of Institutional Reforms
The following sections examine how successful these initiatives—multiparty democracy,
decentralization and community-driven development—have been in instigating economic
development and democratizing institutions, with an eye on changes in social and political norms
of behavior. After a brief look at the post-war macroeconomic situation, Section 5.1 explores
three areas of public goods provision: the impact of decentralization on access to services and
supervision of field staff; how electoral pressures and ethnic-party allegiances influence the
allocation of public spending across constituencies; and how effective CDD has been in
delivering small scale public goods to communities. Section 5.2 then turns to participation,
discussing the impact of the post-war reforms on information flows between citizens and their
elected officials as well as on the direct participation of individuals in community-level
decisions.
5.1 Progress toward Economic Development
Post-war recovery on the macroeconomic level has been relatively rapid and robust. The
Ministry of Finance estimates that the economy has been growing at roughly five percent per
year since the end of the war. Efforts to rebuild public infrastructure have largely restored the
pre-war (although still woefully insufficient) stock of key public goods and flow of basic
services. While this progress is encouraging, the country remains extremely poor: it again fell to
the bottom of the latest Human Development Index, ranking 180th out of the 182 countries
included (United Nations 2009).
While we have no rigorous test of whether local or centralized service provision performs better,
evidence suggests that decentralization has been consistent with improvements in public services
on the ground. Foster and Glennerster (2009) note that access to public services has improved
over the period 2005 to 2007: the percentage of households reporting access to a primary school
within half an hour increased by 5.5 percentage points; to a health clinic by 4.4 ppts; to a
motorable road by 7.3 ppts; to a market by 13.0 ppts (within one hour); and to drinking water by
12.4 ppts (within 15 minutes). The only significant reduction was in access to an agricultural
extension officer, which fell by 5.5 percentage points. For the sector that devolved the most—
13
health—a panel survey of clinics reveals that nearly all indicators of health care quality
(including clinic staffing, supplies and equipment) improved significantly between 2005 and
2008. They further provide suggestive evidence that the gains from decentralization were largest
where the reductions in the distance to power (i.e. from the national capitol to the relevant
district capitol) were greatest. This implies that service improvements were most pronounced in
areas located close to a new district government headquarters but far from Freetown. As a
robustness check, they find no such differential change for education outcomes, which
experienced no decentralization during this time period.
Despite the theoretical gains in supervision, evidence suggests that the newly elected Local
Councillors were not particularly active in overseeing services in the field. Focusing on
healthcare, while Councillors managed to visit roughly half of all clinics during their first year of
office, these visits fell by 50% over the subsequent two years. Furthermore, interviews with
clinic staff suggest that the real gains in oversight and technical assistance occurred within the
ministry itself, with authority devolving from central government bureaucrats to district-level
members of the health management teams. Taken together, these findings suggest that the
substantial improvements in health care may be due more to deconcentration within the ministry
than the oversight of local politicians driven by electoral concerns.
A key sector for economic development, agriculture, remains one of the areas where the
decentralization process is yet to establish a clear division of roles between local councils and the
central government (Casey, 2009). Market construction represents one of the major items of
spending of local councils discretionary funding. Yet, the lack of a central agency devoted to
market development5 and the absence of specific transfers to local councils in this area are
among the factors limiting access to markets for agricultural producers in the country (Srivastava
and Larizza, 2011). One of the authors is currently leading a research team exploring potential
innovative solutions to improve farmers' access to product market, credit market and agricultural
extension. Results from the 2010 Agricultural Household Tracking Survey, the first post-war
nationally representative agricultural household survey, informed some of these pilots. In one of
the projects the team is partnering with the Ministry of Agriculture, Forestry and Food Security
5 The Sierra Leone Produce Marketing Board was dismantled in 1989.
14
(MAFFS) to test the impact of an inventory credit scheme. By providing loans to palm oil
producers, who can use stored product as collateral, the pilot aims to satisfy immediate liquidity
shortages that farmers may face at harvest time. Such cash needs are reported to be one of the
main reasons most sales occur in the peak harvest period, a time when farm-gate prices are
typically low. Evaluation of the pilot will assess the extent to which improved access to credit
may change storage decisions and the timing of product sales. Another project combines
agricultural extension with temporary subsidies to promote the adoption of improved varieties
among rice producers.
One area where electoral politics, service provision and ethnicity come together is in the
allocation of public resources across constituencies. While we have seen that ethnic diversity
does not affect public goods provision within communities, it does appear to influence how much
political patronage external agents bestow upon their constituents. Since ethnic diversity signals
greater political competitiveness, Casey (2012) argues that it attracts greater investment from
candidates on the campaign trail and from politicians once in office. Given the long standing ties
between particular ethnic groups and political parties, she estimates the competitiveness of a
given race for a seat in Parliament or one of the Local Councils based on the ethnic composition
of its constituency. To illustrate: the SLPP and APC have an equal chance of winning a
maximally competitive constituency that has 50% Mende and 50% Temne residents; while the
SLPP has little chance of winning a constituency that is 90% Temne. This logic provides a novel
empirical test of the classic swing voter proposition that both parties favor constituencies with
the weakest underlying party preference (Lindbeck and Weibull 1987). A test based on ethnic
composition—which is largely stable over time and not subject to short term shifts in response to
political patronage—does not suffer the endogenity bias that plagues previous tests based on
reported voting choices.6 Under this framework, Casey finds evidence that moving from a
jurisdiction that is perfectly homogenous to one that is maximally competitive results in a 1.02
standard deviation unit increase in the bundle of campaign goods distributed by national
candidates and $19,577 more public spending by elected local politicians. In a two group
setting, these findings are equivalent to saying that political competition and thus patronage are
6 Voting choices in part reflect transfers from political parties, which arise endogenously from the strategic game played by parties seeking to win elections. See Larcinese, Snyder and Testa (2008) for estimates of the resulting bias.
15
increasing in ethnic diversity, which could provide an alternative explanation for the counter-
intuitive findings regarding clinic quality discussed in Section 3.2.7
Moving down to the community level, decentralizing service provision to villages through CDD
proved effective in delivering smaller scale public goods. Casey, Glennerster and Miguel
(forthcoming) find that strong implementation performance by the GoBifo project improved the
stock and quality of local public goods as well as enhanced general economic welfare in
treatment communities. In particular, the project successfully established Village Development
Committees, plans and community bank accounts; and facilitated greater interaction between
villages and elected local politicians and chiefdom officials. Community verification of project
financial receipts further attests to minimal leakage of project resources, which is no small feat in
an environment of endemic corruption. It also appears that communities used these grants
productively: again using a mean effects approach, GoBifo led to a 0.204 standard deviation unit
increase in the average stock and quality of local public goods (including the presence of a
functional traditional midwife post, latrine and community center, as well as better quality
construction of primary schools and grain drying floors, among other outcomes); and GoBifo
enhanced general economic welfare by 0.376 standard deviation units (including the presence of
petty traders, number of goods on sale, household assets and amenities, among others). These
results suggest that CDD is a reasonable approach to delivering small scale local public goods in
a way that is equitable, accountable and low cost.
5.2 Progress toward Democratizing Institutions
Our research suggests that the post-war institutional reforms have increased information flows
between citizens and local politicians, with positive impacts on voting behavior and oversight of
local public goods. Yet we find little evidence to suggest that community-level interventions
have led to fundamental changes in social norms, particularly with respect to the voice and
participation of women and youths. 7 Note that of all the measures considered by Glennerster et al, clinic quality is the least influenced by local collective action, where major policy decisions—including clinic location choice, procurement and hiring/firing of medical staff—are all controlled more centrally. Health clinic quality may thus provide another example of how winning politicians reward more diverse and politically competitive jurisdictions with better public goods once in office.
16
Decentralization has lived up to its purported informational advantages in terms of increasing
how much voters know about their politicians, which in turn reduces the salience of ethnic
identity in voting choices and campaign spending. As decentralization brings government closer
to the people, Casey notes that voters have more information about local politicians (for
example, while 37% of respondents could correctly name their Local Councillor, only 17% could
name their Parliamentarian) and more opportunities to interact with them (while 52% of
communities reported being visited by their elected Councillor in the past year, only 27% report
a visit from their MP). Adding a layer of information asymmetry onto the standard swing voter
investment model, Casey demonstrates that increasing the amount of information voters have
about candidates (beyond party affiliation), makes voters more likely to cross traditional ethnic-
party lines for particularly attractive rival party candidates. Exploiting the information
differences across levels of government and using individual fixed effects, she shows that the
same voters are 11.3 percentage points more likely to cross traditional ethnic group-political
party allegiances when voting in local as opposed to national elections. Knowing this, parties in
turn respond by de-emphasizing ethnic-party allegiances in allocating campaign spending across
constituencies. While parties continue to favor more diverse constituencies for both local and
national races, the spending in local races is roughly half as responsive to ethnic composition as
that for national elections. These findings speculatively suggest that providing voters with better
information about candidates could help diffuse the ethnic tensions surrounding elections.
While CDD strengthened village-level public good provision and created meaningful links
between villagers and the lowest tiers of elected government, it did not fundamentally influence
social norms nor communal capacity for collective action. Yet, taking a step back, let us first
question the assumption that the war left the communities highly compromised in terms of social
cohesion and their ability to work together. Note that baseline levels of cohesion were high: over
81% of respondents had trusted a neighbor with goods to sell on their behalf in the local market;
the average person was a member of more than 2 of 5 common social groups; and only 22%
reported having a financial conflict in the preceding year. Glennester et al. further show that
there is no evidence to negatively link ethnic diversity with such cohesion and the resulting
ability to act collectively. In addition, Bellows and Miguel emphasize that if anything, survivors
17
of violence emerged from the war even more politically and socially active than they were
before. Such initially high levels of cohesion throw into doubt the necessity of the social
facilitation aspect of CDD in this context.
That said, there were clear social divisions—between women and men, and youths and their
elders—that were creating frustration, and chronic poverty suggests that greater collective action
toward local development would likely be welfare enhancing. Yet while CDD explicitly targeted
these areas for remedy, we find no evidence that it had any lingering impact on the voice and
participation of women and youth, nor on the likelihood that communities were able to take up
development opportunities arising after the project ended. While these null results held true
across a wide variety of outcomes, the following two examples from our “structured community
activities” (SCAs) are illustrative. The SCAs were designed to unobtrusively observe members
of treatment and control communities engaged in concrete, real-world collective activities and
decisions. Regarding voice, the research team found no difference in the number of women who
spoke publicly in a community meeting to decide between two small gifts offered by the research
team as a thank you for their participation in the study. Regarding collective action, exactly 62
treatment and 64 control villages (roughly 50% across the board) took up a voucher opportunity
to purchase building materials at a subsidized price for use in a community project. These
findings (along with the other indicators studied) suggest quite conclusively that CDD had no
impact on underlying attitudes towards women and youths, and did not serve as a catalyst for
collective action beyond the life of the project.
6. Concluding Remarks
Many would have said that Sierra Leone was condemned to be a basket case because of its war
history, ethnic diversity and poverty, but the economic and political progress of the last decade
belies this. In fact, war legacies and ethnic diversity do not appear to necessarily hinder local
collective action. And while national politics remain heavily ethnic, power transferred
peacefully between parties, and voters are willing to cross traditional ethnic allegiances when
they have better information. Turning to public goods, decentralization has been compatible
with steady improvements in service delivery, greater interaction between citizens and their
18
elected representatives, and enhanced supervision of front line workers by district-level
managers. Ten years ago, few observers would have thought this last decade of peace and
prosperity was possible for Sierra Leone.
In terms of policy lessons, it is useful to consider the role that external actors played in these
achievements. Taken together, our research suggests that external assistance and interventions to
support large scale institutional reforms have met with greater success than those targeting
community level norms and dynamics. In particular, international donors and foreign
governments made positive contributions to ending the civil war, restoring multiparty democracy
and decentralizing public services. Implemented in partnership with the Government of Sierra
Leone, these reforms have translated into substantial, and quite tangible, benefits on the ground.
The experience with CDD, however, suggests that fundamentally altering community-level
hierarchies and social norms is incredibly difficult, and not something that we, as outsiders, yet
know how to do effectively.
Moving forward, much uncertainty remains regarding the potential for sustained economic
growth and institutional development in Sierra Leone. The discovery of large deposits of natural
resources (including iron ore and oil) provides hope for improved incomes but may also place
additional pressure on weak institutions. As Sierra Leone celebrated 50 years of independence in
2011, the country faced an open question of whether these newly found resources would
contribute more to general development than diamonds have to date. As with the successes we
have seen in the last several years, the country’s further progress ultimately rests on the
resilience and determination of Sierra Leoneans striving for a better future.
19
References
Acemoglu, D., S. Johnson and J.A. Robinson (2001) “The Colonial Origins of Comparative Development: An Empirical Investigation,” American Economic Review, 91(5), 1369-1401.
Alesina, Alberto, Arnaud Devleeschauwer, William Easterly, Sergio Kurlat and Romain Wacziarg (2003) “Fractionalization,” Journal of Economic Growth, 8(2), 155-194.
Alesina, Alberto, Reza Baqir, and William Easterly (1999) “Public Goods and Ethnic Divisions,” Quarterly Journal of Economics, 114(4), 1243-1284.
Banerjee, Abhijit and Lakshmi Iyer (2005) “History, Institutions, and Economic Performance: The Legacy of Colonial Land Tenure Systems in India.” American Economic Review, 95(4): 1190-1213.
Bardhan, Mookherjee (2006) “Decentralisation and Accountability in Infrastructure Delivery in Developing Countries,” The Economic Journal, 116(508), 101-127.
Bellows, John and Edward Miguel (2009) “War and Local Collective Action in Sierra Leone,” Journal of Public Economics, 93(11-12), 1144-1157.
Casey, Katherine (2009), “Decentralization in Practice” in Decentralization, Democracy, and Development: Recent Experience from Sierra Leone (Ed. Yongmei Zhou) Washington D.C.: World Bank.
Casey, Katherine (2012) “Crossing Party Lines: The Effects of Information on Redistributive Politics,” Stanford GSB Working Paper No. 2099.
Casey, Katherine, Rachel Glennerster and Edward Miguel (forthcoming) “Reshaping Institutions: Evidence on Aid Impacts Using a Pre-Analysis Plan,” Quarterly Journal of Economics.
Dongier, Philippe, Julie Van Domelen, Elinor Ostrom, Andrea Rizvi, Wendy Wakeman, Anthony Bebbington, Sabina Alkire, Talib Esmail and Margaret Polski (2003) “Chapter 9: Community-Driven Development.” In The Poverty Reduction Strategy Sourcebook Volume 1, 301-331.Washington DC: The World Bank.
Easterly, William and Ross Levine (1997) “Africa’s Growth Tragedy: Policies and Ethnic
Divisions,” Quarterly Journal of Economics, 112, 1203-1250.
Engerman, Stanley L. and Kenneth L. Sokolof (1997) “Factor Endowments, Institutions, and Differential Paths of Growth Among New World Economies: A View from Economic Historians of the United States.” In How Latin America Fell Behind. Stanford: Stanford University Press.
20
Fearon, James D. (2003) “Ethnic and Cultural Diversity by Country,” Journal of Economic Growth, 8(2), 195-222.
Foster, Elizabeth and Rachel Glennerster (2009) “Impact of Decentralization on Public Services: Evidence to Date” in Decentralization, Democracy, and Development: Recent Experience from Sierra Leone (Ed. Yongmei Zhou) Washington D.C.: World Bank.
Freedom House, Inc. (2009) “Country Report: Sierra Leone,” accessed 8 July 2010 on http://www.freedomhouse.org/template.cfm?page=22&year=2009&country=7699.
Glennerster, R., E. Miguel and A. Rothenberg (2010) “Collective Action in Diverse Sierra Leone Communities” NBER Working Paper No. 16196.
Habyarimana, J., M. Humphreys, D. Posner, and J. Weinstein (2007).“Why Does Ethnic Diversity Undermine Public Goods Provision?” American Political Science Review, 101(4), 709-725.
Habyarimana, J., M. Humphreys, D. Posner, and J. Weinstein (2009) Coethnicity: Diversity and the Dilemmas of Collective Action. Russell Sage: New York.
Human Rights Watch (1999) Sierra Leone: Getting Away with Murder, Mutilation, and Rape. New York: Human Rights Watch.
Humphries, Macartan, Weinstein, Jeremy M. (2006) “Handling and Manhandling Civilians in Civil War: Determinants of the Strategies of Warring Factions,” American Political Science Review 100(3), 429-447.
Kling, J., J. Liebman and L. Katz (2007) “Experimental Analysis of Neighborhood Effects,” Econometrica, 75(1), 83-119.
Keen, David (2005) Conflict and Collusion in Sierra Leone. London: James Currey; New York: Palgrave.
Lacrincese, Valentino, James M. Snyder and Cecilia Testa (2008) “Testing Models of Distributive Politics Using Exit Polls to Measure Voter Preferences and Partisanship,” manuscript.
Lindbeck, Assar and Jorgen W. Weibull (1987) “Balanced-budget redistribution as the outcome of political competition,” Public Choice, 52, 273-297.
Mansuri, Ghazala and Vijaynedra Rao (2004) “Community-Based and -Driven Development: A Critical Review.” World Bank Research Observer, 19(1): 1-39.
Mansuri, Ghazala and Vijayendra Rao (2012) Localizing Development: Does Participation Work? Washington, DC: The World Bank.
21
Miguel, Edward (2004) “Tribe or Nation? Nation Building and Public Goods in Kenya versus Tanzania,” World Politics 56(3), 327-362.
Miguel, Edward (2009) Africa’s Turn? Cambridge, MA: The MIT Press.
Miguel, Edward and Mary Kay Gugerty (2005) “Ethnic diversity, social sanctions, and public goods in Kenya,” Journal of Public Economics, 89(11-12), 2325-2368.
Ngugi wa Thiong’o (2009) Something Torn and New: An African Renaissance. NY: Nation Book.
Oates, Wallace E. (1999) “An Essay on Fiscal Federalism,” Journal of Economic Literature, 37(3), 1120-1149.
Reno, William (1995) Corruption and State Politics in Sierra Leone. Cambridge and New York: Cambridge University Press.
Richards, Paul (1996) Fighting for the Rainforest: War, Youth and Resources in Sierra Leone. London: James Currey; Portsmouth, NH: Heinemann for the International African Institute.
Srivastava, Vivek and Marco Larizza, (2011), “Decentralization in Postconflict Sierra Leone: The Genie is out of the Bottle” in Yes Africa Can: Success Stories from a Dynamic Continent (Eds. Putnam Chuhan-Pole and Manka Angwafo) Washington D.C. :World Bank.
United Nations (1993) Human Development Report 1993. New York: United Nations Development Program, Oxford University Press.
United Nations (2004) Human Development Report 2004. New York: United Nations Development Program, Oxford University Press.
United Nations. (2009) Human Development Report 2009. New York: United Nations Development Program, Oxford University Press.
World Bank (2007) “World Development Report 2008: Agriculture for Development.” Washington, D.C.: World Bank.
22
Figure 1
Trends in GDP per capita 1960-2010 (in constant 2000 US dollars) for sub-Saharan Africa
and Sierra Leone
0
100
200
300
400
500
600
700
1960 1970 1980 1990 2000 2010
Sub-Saharan Africa
Sierra Leone
GD
P p
er c
apit
a(c
on
stan
t 2
00
0 U
S$)
Notes on figure: i) source of data is World Development Indicators, The World Bank; and ii) the sub-Saharan African mean is the “all income levels” series.